In this paper we present a linguistic approach to the analysis of proper names. The basic assumption of our approach is that proper names are linguistic units of text that should be treated using the same methods that are applied to text in its totality. We illustrate the inflectional and derivational properties of simple and multi-word proper names on the example of Serbian, and describe how these properties have been formalized in order to develop e-dictionaries of the DELA type. In order to support multi-lingual applications we have developed a model of a multilingual relational dictionary of proper names based on an ontology, as well as an actual database. Finally, we outline how the developed dictionaries and database can be used in real monolingual and multi-lingual applications, such as information extraction.
[1]
Denis Maurel,et al.
Multilingual Ontology of Proper Names
,
2005
.
[2]
M. Silberztein,et al.
Dictionnaires électroniques du français
,
1990
.
[3]
Duško Vitas,et al.
Regular Derivation and Synonymy in an E-Dictionary of Serbian
,
2005
.
[4]
Duško Vitas,et al.
Using Textual and Lexical Resources in Developing Serbian Wordnet
,
2004
.
[5]
Denis Maurel,et al.
"Beijing Frowns and Washington Pays Close Attention" Computer Processing of Relations between Geographical Proper Names in Foreign Affairs
,
2000,
NLDB.
[6]
Duško Vitas.
Morphologie dérivationnelle et mots simples: Le cas du serbo-croate
,
2004
.